The overall goal of this research is to better understand the contribution of the cerebellum and cerebrocerebellar networks to cognitive function by combining functional MRI (fMRI) and transcranial magnetic stimulation (TMS) techniques with tasks of verbal working memory. The Sternberg verbal working memory task, with three temporally distinct phases of encoding, maintenance, and retrieval, is an ideal model system for studying cerebro-cerebellar contributions to cognition. Our studies to date have supported a cerebrocerebellar verbal working memory model that emphasizes a functional link between neocortical articulatory control regions, such as Broca's Area, to the superior cerebellum, and a second network that links neocortical substrates of phonological storage, located in inferior parietal regions, to the inferior cerebellum. We propose to further test and extend this model in three specific aims. The first will emphasize the encoding phase and will test the hypothesis that the superior cerebellum and posterior frontal regions participate in the initial preparation of the articulatory trajectory independent of stimulus modality. The second aim will emphasize the maintenance phase. We will test the hypothesis that the right inferior cerebellum is functionally linked to the left inferior parietal region and that activation in these regions will be modulated in tandem when phonological storage demands are systematically manipulated. In addition, we will test the hypothesis that a transient phonological storage deficit syndrome, similar to that observed in left temporal-parietal lesioned patients, can be reproduced in the TMS lab using stimulation to left inferior parietal and right inferior cerebellar regions. The third aim will emphasize the retrieval phase. The retrieval phase of the Sternberg task, which requires utilization of the stored information, has elicited the least amount of cerebellar activation in our event-related investigations. We hypothesize that by increasing the executive demands of the task - by imposing additional processing requirements - we will observe increases in retrieval-phase-related cerebellar activation. We further hypothesize that the amount of behavioral disruption produced by TMS in the retrieval phase will be correlated with the amount of retrieval-phase related cerebellar activation.